This is just something that popped into my head after watching a couple YouTube tutorials. And I’m genuinely confused about the differences between the two.
Bump/Normal Maps: What's The Difference?
Cadedra
#2
A few tips:
The result of using both types of maps is that the interpretation of normal direction over the surface is altered, so object appears to have bumps when it actually does not have. That said, none of the two does an actual change on geometry
Bump maps:
1-Takes a BW map to define which parts will look as “more displaced” and which parts “less displaced” or “not at all”
2-Easy to create/modify. You can even do it manually, using PhotoShop or something like that.
3-Computationally more expensive. Somehow the computer has to go through a process that turns bump map information into something very similar to a normal map in the end, if not exactly equal. A pain for real time, although it is hard to feel the difference in production renderings.
4-Since most of the time you have a mere 0 to 255 grayscale it is very hard to get subtle slopes, as it would take a gradient that rises up one unit every… a maybe long distance. You can end with a “stair look”.
5-Bump maps can also help generating another kind of maps depending of the case. I have many times used Bump map data to aid generation of reflection maps, color maps, etc, because it is not uncommon that in some materials bump behavior might be somehow related to the behavior of some other properties.
Normal maps:
1-Takes a color map whose RGB values define local vectors that tells which direction the surface “is facing at” at some point.
2-Hard to create/modify. Most of the time you need dedicated software. The behavior of the map is not intuitive just by looking at it, so you have to wait and see the final output. And do not even dream on mixing normal maps without knowing the right procedure first, or you end up with a wrong look. So maybe put your money on additional software and training.
3-Computationally cheap. The computer is getting a straight data about normal behavior that can be used right away for shading. No wonder it is widely used for games.
4-You can get any slopes, since what normal maps define is which direction the surface is oriented to, and it can be subtle.
5-Normal maps are the lonely island. They will (almost?) never aid for other map’s generation or viceversa.
Cadedra summed it up nicely.
Normal maps use direction rather than intensity to dictate the 3D surface.
Normal maps can be a lot more accurate but cant be created manually using photoshop or other editing techniques, so much harder to edit.